Clinician Communication About Uncertainty During Parkinson Disease Diagnostic Consultations Authors: Marij A. Hillen, Elisabeth Kurpershoek, Mark H.B. Huisman, C. Machiel Pleizier, Annemarie Vlaar, Jeroen Blankevoort, Evelien Zoons, Dareia S. Roos, Marianne de Visser, Rob M.A. de Bie, Joke M. Dijk Published: 2026 ABSTRACT Background and ObjectivesParkinson disease (PD) entails much uncertainty, regarding diagnosis, disease course, and treatment. How clinicians communicate about such uncertainty may affect patients' understanding of the information and their well-being. However, clinicians rarely receive formal training in how to optimally discuss uncertainty during diagnostic consultations. We assessed (1) how clinicians communicate about uncertainty, (2) how their communication about uncertainty predicts patients' feelings of uncertainty and trust, and (3) the moderating effect of clinicians' and patients' tolerance for uncertainty.MethodsWe recorded diagnostic consultations regarding PD (initial and follow-up within 6 weeks) for 50 patients. The frequencies of clinicians' verbal expressions of uncertainty were systematically coded for topic, initiative, explicitness, and communication approach. Moreover, clinicians' reactions on patients' and companions' uncertainty expressions were coded. Patients and clinicians completed multiple surveys assessing background characteristics, uncertainty tolerance, feelings of uncertainty, and trust. We conducted descriptive analyses and fitted mixed linear models.ResultsClinicians expressed uncertainty on average 20 times over both diagnostic consultations combined (range 6–67; 0.40 p/min). Most statements concerned medication (31%) and diagnosis (30%), and uncertainty was predominately conveyed implicitly (mean = 74%). The most frequently observed communication approach (30%) was outlining a plan to obtain more certainty. Patients or companions expressed uncertainty on average 12 times (range 1–29; 0.20 p/min), to which clinicians most often responded by providing more information (67%). Patients' feelings of uncertainty decreased after the diagnostic consultation(s). Clinicians' uncertainty tolerance was not associated with their uncertainty communication. Neither patients' nor clinicians' uncertainty tolerance moderated the association between uncertainty communication and patient-reported outcomes.DiscussionUncertainty is integral to PD diagnostic consultations and does not seem to have harmful effects on patient well-being. Its discussion often remains implicit and is strongly focused on justifying or reducing the uncertainty. More systematic research could assess the effects of various ways to convey uncertainty on patients' well-being and trust. Introduction Parkinson disease (PD) is a progressive neurodegenerative movement disorder, leading to reduced quality of life and increased risk of frailty.1,2 Receiving a PD diagnosis is distressing and has a lasting impact on patients.3 It is important to note that PD is fraught with multiple uncertainties.4 First, the accuracy of the diagnosis is uncertain: a meta-analysis in 2016 demonstrated a diagnostic accuracy of 80.6%.5 This is due to the difficulty in differentiating PD from atypical parkinsonian disorders, particularly early in the disease.6 Second, PD progression is very heterogeneous. Motor and nonmotor symptoms vary considerably between individuals, and the speed of progression is highly variable. Making personalized predictions about symptom progression and its impact on patients' lives is challenging.6 Third, the effects of treatments are uncertain. Disease progression cannot be stopped or reversed, yet symptoms can be reduced by dopamine supplementation therapy.7 The long-term efficacy and side effects of dopaminergic pharmacotherapy vary widely between people. Moreover, owing to progressive degeneration of dopaminergic neurons, medication regimen alterations are required over time and bothersome side effects occur frequently.8 Despite these uncertainties, clinicians are expected to provide accurate and realistic information to patients about diagnosis and its consequences, without causing undue distress or reducing patient trust or satisfaction. Effectively fulfilling this task is crucial because diagnostic communication affects patients' long-term quality of life and satisfaction with care.9,10 Significant proportions of people with PD retrospectively report dissatisfaction with diagnostic information.9,11,12 Patients' suboptimal diagnostic experiences may contribute to reduced treatment compliance, as well as impaired clinician-patient relationships and patient well-being.13,14 Neurologists generally believe that patients deserve to be informed about the uncertainty surrounding PD at diagnosis, yet find this challenging.15 Clinicians' struggle and anxiety to discuss may be rooted in medical education, which strongly emphasizes a quest for certainty, whereas acknowledging uncertainty is generally not rewarded.16,17 Clinicians may fear that conveying uncertainty toward patients may harm their reputation or patients' trust.18-20 Currently, practical guidance and training for clinicians on optimally discussing uncertainty with patients is scarce.21 This is problematic because evidence suggests that how clinicians communicate uncertainty may determine its effects on patients' satisfaction, trust, and anxiety.18,19,21-24 For example, patients' trust may be affected by how explicitly or implicitly a clinician conveys uncertainty, which may ultimately affect health behaviors such as adherence to the clinician's advice.25 Three key aspects of clinicians' uncertainty communication are presumed to particularly affect patients: the frequency of these expressions, their level of explicitness, and the degree of emotion-focused support conveyed to manage the uncertainty.21 Empirical evidence is needed to develop optimal communication strategies for communicating uncertainty during PD diagnosis conversations. A personal characteristic that may crucially affect clinicians' communication about uncertainty and how it is perceived by patients is “tolerance for uncertainty”—that is, a person's general ways of responding to uncertainty.26 Evidence indicates that lower tolerance for uncertainty in clinicians is associated with more fear of making mistakes, overuse or underuse of diagnostic testing, and burnout.27,28 For patients, preliminary evidence suggests that their uncertainty tolerance determines how they are affected by uncertain information.27 Thus, clinicians' uncertainty tolerance may determine how they discuss uncertainty with patients, whereas patients' uncertainty tolerance may moderate how they are affected by such communication.27 Yet, these relationships have never been empirically assessed in PD. We aimed to (1) assess clinicians' communication about uncertainty, (2) test how uncertainty communication predicts patients' feelings of uncertainty and trust in PD diagnostic consultations, and (3) test how both clinicians' and patients' tolerance for uncertainty moderates associations between uncertainty communication and patient outcomes. Methods Design This analysis was part of a larger prospective observational mixed-methods longitudinal study on information provision regarding PD diagnosis.29 Quantitative data (questionnaires and systematic coding of video-recorded clinician-patient interactions) and qualitative data (interviews) were used to investigate clinician communication and patient-reported outcomes regarding PD diagnostic consultations. Diagnostic consultations were defined as the first and any second consultation within a six-week period. In this analysis, we used a subset of the collected data. Participants Participants were clinicians, patients, and patients' companions. Patients received a PD diagnosis during the video-recorded consultations at one of 6 participating neurology outpatient clinics in the Netherlands. We excluded patients who had insufficient Dutch language proficiency, were previously evaluated for suspected PD, had other unrelated severe medical conditions or disabilities, or had preexisting cognitive impairment. Patients who had received a different diagnosis than PD during the recorded consultations, based on the International Parkinson and Movement Disorder Society criteria, were also excluded.30 Procedures Study procedures are displayed in Figure 1. Eligible patients were identified by neurologists, who scanned general practitioners' referral letters for suspected PD. Eligible patients received study information. The research was introduced in general terms, to prevent intervening on the diagnostic process. Before the first consultation (T0), participating patients and their clinicians completed a baseline questionnaire (on paper or digitally, depending on preference), assessing sociodemographic and psychological characteristics, and signed written informed consent. Consultations were video recorded using a secure, offline iPhone. Up to 2 weeks afterward, patients completed a second questionnaire (T1). If a second diagnostic consultation was planned within 6 weeks after the first, this was also video recorded, followed by another patient questionnaire within 2 weeks (T2). Patients completed a final questionnaire 6 to 8 weeks after their last consultation (T3). Study Procedures Measures Systematic Coding of Communication We adhered to the reporting guidelines for systematic coding of observed human behavior (SCOBe; eAppendix 1). Clinicians', patients', and companions' communication about uncertainty was coded in video-recorded diagnostic consultations using observational coding software (Noldus The Observer XT 15, Wageningen, the Netherlands). The coding manual was adapted from existing coding instruments developed and applied in other medical contexts.31-33 Modifications were made based on study team discussions and a scoping review.21 The modified coding instrument was extensively tested, refined, and then finalized before application. We excluded any parts of the consultation that involved small-talk or physical examination, coding only “active consultation.” The uncertainty communication coding instrument is displayed in eAppendix 2. In summary, each instance in which the clinician expressed uncertainty was identified and coded for:Topic, i.e., the subject of uncertainty (5 main categories).Initiative, i.e., who initiated the uncertainty expression (clinician, patient, or companion).Explicitness, i.e., whether the uncertainty was conveyed using explicit statements (e.g., “I don”t know” or “it is unclear”) or implicitly, using more subtle phrasing including adverbs such as “maybe” or modal verbs such as “might”.34Communication approach, i.e., which method(s) the clinician used to convey the uncertainty (16 options organized into 4 main categories: information-focused, emotion-focused, other, or none; 1 or 2 approaches selected per expression). Topic, i.e., the subject of uncertainty (5 main categories). Topic, i.e., the subject of uncertainty (5 main categories). Initiative, i.e., who initiated the uncertainty expression (clinician, patient, or companion). Initiative, i.e., who initiated the uncertainty expression (clinician, patient, or companion). Explicitness, i.e., whether the uncertainty was conveyed using explicit statements (e.g., “I don”t know” or “it is unclear”) or implicitly, using more subtle phrasing including adverbs such as “maybe” or modal verbs such as “might”.34 Explicitness, i.e., whether the uncertainty was conveyed using explicit statements (e.g., “I don”t know” or “it is unclear”) or implicitly, using more subtle phrasing including adverbs such as “maybe” or modal verbs such as “might”.34 Communication approach, i.e., which method(s) the clinician used to convey the uncertainty (16 options organized into 4 main categories: information-focused, emotion-focused, other, or none; 1 or 2 approaches selected per expression). Communication approach, i.e., which method(s) the clinician used to convey the uncertainty (16 options organized into 4 main categories: information-focused, emotion-focused, other, or none; 1 or 2 approaches selected per expression). Each expression of uncertainty by the patient or companion was also flagged and coded for the clinician's initial response (6 options; e.g., “minimal recognition”; eAppendix 2). All consultations were independently coded in pairs of 2 out of 6 observers (MAH with L.B., J.B., M.B., E.C., or E.K.). All coders were trained through collaborative coding supervised by the coding instrument developers (E.K. and M.H.). Coders held consensus meetings to compare, resolve discrepancies, and finalize coding choices. Cohen Kappa was calculated to assess inter-rater reliability. Global agreement was defined as agreement regarding the occurrence and timing of uncertainty expressions and their topic (for clinician expressions), or uncertainty expressions and clinician's response (for patient/companion expressions). Detailed agreement pertained to agreement on the occurrence and timing of all coded aspects (i.e., for clinician expressions: topic, initiative, explicitness, and communication approach and for patient expressions: clinicians' response). Average detailed agreement for the first 5 consultations in every coding pair ranged between κ = 0.09 and 0.48, indicating insufficient agreement. We, therefore, decided to continue coding all video-recorded consultations with 2 independent coders and held consensus meetings to ensure coding reliability. Questionnaires Patients Sociodemographic characteristics assessed at T0 were age, sex, educational level, ethnicity, and mother tongue. Uncertainty tolerance (T0) was measured using the validated Dutch version of the 12-item Intolerance of Uncertainty Scale short form (IUS-12).35,36 Higher scores indicate lower tolerance of uncertainty. Internal consistency was high at α = 0.87. Feelings of uncertainty (T0, T1, T2, T3) were assessed using a selection of 10 of the 30 items from the MIshel37 Uncertainty of Illness Scale (MUIS) that were relevant for the current setting. Higher scores indicate higher uncertainty. Internal consistency at the 4 measurement points ranged from α = 0.69 (acceptable) to 0.82 (good). Trust in clinician (T1, T2, T3) was assessed using the validated 5-item Trust in Oncologist Scale short form (TiOS-sf), with phrasing adapted slightly to the neurology setting.38 Higher scores indicate higher levels of trust. Internal consistency ranged from α = 0.63 (acceptable) to α = 0.86 (good). IUS-12, MUIS, and TiOS-sf items were all answered on a 5-point Likert scale (1 = totally disagree to 5 = totally agree). Cognitive functioning was assessed by trained researchers using the Parkinson's Disease Cognitive Rating Scale (PDCRS), between 3 and 24 months after diagnosis. Scores were dichotomized as follows: 1 = no cognitive impairment (scores >80) or 2 = indication of mild cognitive impairment (MCI) (scores 64–81).39 Clinicians Sociodemographic and professional characteristics assessed were age, sex, and years of professional experience (T0). Professional uncertainty tolerance (T0) was measured using the Physicians' Reactions to Uncertainty (PRU) scale.40,41 The PRU contains 2 subscales: “anxiety due to uncertainty” and “reluctance to disclose uncertainty to patients.” Items were answered on a 6-point Likert scale (1 = totally disagree to 6 = totally agree). Higher mean scores indicate lower tolerance. Internal consistency was good at α = 0.83 overall (0.82 and 0.73 for reluctance and anxiety subscales, respectively). Hindrance from the video camera equipment (T1) was assessed using 1 item, reported on a visual analog scale (VAS; 0 = “not bothered” to 100 = “extremely bothered”). Analyses We extracted descriptive data on patients', companions', and clinicians' observed uncertainty communication from the observational software program. Regarding clinicians' uncertainty expressions, we calculated topics, initiative, explicitness, and emotion-focused communication approaches for the statistical analysis. Frequency of topics was described in both absolute (number of uncertainty expressions per consultation) and relative (average number of uncertainty expressions per minute of active consultation) terms. Initiative was defined as the percentage of all uncertainty expressions that were initiated by the clinician, the patient, or the companion. Explicitness was defined as the percentage of all uncertainty expressions that were explicit (vs. implicit). An emotion-focused communication approach was defined as the percentage of all communication approaches used when expressing uncertainty, in which clinicians used one of 6 emotion-focused approaches. Finally, we calculated the frequency of patient and companion expressions of uncertainty and the percentages of 6 possible initial responses by clinicians. Descriptive statistics (i.e., mean/median, SD/interquartile range, range, percentage) were used to describe the sample. For aim 1 (i.e., to describe uncertainty communication), we calculated the following: (1) percentages for clinician responses to patient/companion uncertainty expressions and (2) frequencies (absolute and rate/minute) and percentages for clinician uncertainty expressions, categorized by total expressions and topics, initiative, explicitness, and communication strategies. For aim 2 (i.e., to test associations between clinicians' uncertainty communication and patient uncertainty and trust), we (1) used repeated-measures analyses of variance (ANOVA) to test potential changes in patient uncertainty and trust over time; (2) used multiple regression analyses to test associations between clinicians' uncertainty communication during the first consultation and patients' trust (T1) and uncertainty (change T0-T1); and (3) repeated the models from (2) for communication during the second consultation and trust (T2) and uncertainty (change T1-T2). For aim 3 (i.e., to test the moderating effect of clinicians' and patients' uncertainty management), we (1) planned to add patient uncertainty tolerance as a moderator to regression models and (2) ran univariate linear mixed-effects regressions to explore associations between clinician uncertainty tolerance and their uncertainty communication. Standard Protocol Approvals, Registrations, and Participant Consents Ethical approval was obtained from the local ethics review board of Amsterdam UMC (METC-2017-271). Written informed consent was obtained from all study participants. Data Availability Anonymized questionnaire data are available on request from the authors. Video recordings are not available because informed consent was not obtained to share the data for other purposes than this research. Results Recruitment and Sample Figure 2 shows recruitment and inclusion. We approached 229 patients for study participation, of whom 88 declined participation. Of the 141 patients who consented to participation, 71 received a diagnosis other than PD and 4 were excluded for other reasons. Eventually, 50 patients were included in our analyses. Table 1 lists patient and clinician characteristics. The mean age of patients was 72.1 years (range 51–91), and 32% identified as female. Education level was high for 44%, middle for 30%, and low for 26%. All patients were native Dutch speakers. Based on PD-CRS scores, 18% had MCI, while all others had no cognitive impairment. Consultations were conducted by 28 different clinicians (Mage = 40.9 years, range 30–64), 57% of whom were women. Mean years of experience among residents (n = 8) were 4.8 years (SD = 1.6), and for neurologists (n = 18) were 15.0 years (SD = 7.0) (not assessed among physician assistants). Patients' uncertainty tolerance was 2.9 on average (SD = 0.7, range 1.5–4.4). Clinicians' overall tolerance for uncertainty was 28.1 on average (range 15–39). Clinicians reported minimal hindrance from the video camera (M = 2.2; SD = 2.1). Patient Recruitment and Follow-Up Table 1Sample CharacteristicsPatients (N = 50)Clinicians (N = 28)Age, mean y (SD)72.1 (9.2)40.9 (9.3)Self-identified sex, female, n (%)16 (32)16 (57)Type of clinician, n (%) Neurologist18 (64) Resident8 (29) Physician assistant2 (7)Children, yes, n (%)40 (80)Living situation, alone, n (%)11 (22)Work (incl. voluntary), yes, n (%)15 (3)Education level, n (%) None, primary education13 (26) (Basic) vocational training15 (30) University, higher professional education22 (44)Native language, Dutch, n (%)50 (100)PD-CRS (n = 32), mean score (SD)a95.3 (16.1) Mild cognitive impairment (MCI), n (%)b6 (19)Type of medical center, academic, n (%)13 (26)Tolerance of uncertainty, mean score (SD)c,d2.9 (0.7)28.1 (6.4) Anxiety subscale, mean score (SD)eNA16.3 (4.2) Reluctance subscale, mean score (SD)eNA11.9 (3.0)Abbreviations: MCI = mild cognitive impairment, PD-CRS = Parkinson’s Disease Cognitive Rating Scale; NA = not applicable.aPD-CRS: Parkinson's Disease Cognitive Rating Scale.42 Owing to COVID-19 restrictions at the time of participation, some assessments were not conducted or were delayed up to 24 mo after diagnosis.bPD-CRS scores: scores >80 indicate no cognitive impairment; scores between 64 and 81 indicate mild cognitive impairment (MCI).c12-Item Intolerance of Uncertainty Scale short form (IUS-12).35,36 Potential score range = 1–5. Higher scores indicate lower tolerance of uncertainty.dPhysicians' Reactions to Uncertainty (PRU) scale. Potential range = 10–60. Higher mean scores indicate lower tolerance.ePotential range for each subscale = 5–30. Higher scores indicate lower anxiety and lower reluctance, respectively. Sample Characteristics Abbreviations: MCI = mild cognitive impairment, PD-CRS = Parkinson’s Disease Cognitive Rating Scale; NA = not applicable. PD-CRS: Parkinson's Disease Cognitive Rating Scale.42 Owing to COVID-19 restrictions at the time of participation, some assessments were not conducted or were delayed up to 24 mo after diagnosis. PD-CRS scores: scores >80 indicate no cognitive impairment; scores between 64 and 81 indicate mild cognitive impairment (MCI). 12-Item Intolerance of Uncertainty Scale short form (IUS-12).35,36 Potential score range = 1–5. Higher scores indicate lower tolerance of uncertainty. Physicians' Reactions to Uncertainty (PRU) scale. Potential range = 10–60. Higher mean scores indicate lower tolerance. Potential range for each subscale = 5–30. Higher scores indicate lower anxiety and lower reluctance, respectively. Descriptive Characteristics for Uncertainty Communication The duration of the consultation, excluding physical examination, was on average 30 minutes (SD = 11, range 6–67) for first consultations (n = 50) and 22 minutes (SD = 13, range 5–69) for second consultations (n = 41). Thirty patients (60%) were accompanied by a companion in at least one of their consultations. Table 2 provides descriptive characteristics for uncertainty communication in the recorded consultations. Regarding frequency, clinicians uttered on average 20 uncertainty expressions over both consultations combined (SD = 10, range 1–51), at a rate of 0.4 expressions per minute (SD = 0.1, range 0.2–0.7). During first consultations, most uncertainty expressions pertained to the diagnostic label (mean (M) = 39%; SD = 24%, range 0%–100%), and during second consultations, to medication (M = 44%; SD = 31%, range 0%–100%). Regarding explicitness, on average 26% of uncertainty expressions were explicit over both consultations combined. Information-focused approaches were mostly observed (72%). The most frequent of these was “offering a sense of control, by reducing uncertainty” (25%), for example, “If the [dopamine replacement] therapy improves your symptoms, it will also provide more certainty about whether you have PD.” Emotion-focused approaches constituted 15% of all approaches, the most frequent of which was “providing hope” (12%), for example, “But based on everything we have discussed, I think it is very likely that you will notice positive effects of this medication - I really expect that to happen.” Table 3 provides illustrative examples of all observed communication approaches. Table 2Descriptive Statistics for Observed Uncertainty Communication During Recorded ConsultationsAll consultations (N = 50)First consultation (N = 50)Second consultation (N = 41) Total expression, mean frequency (SD) 20 (10) 12 (5) 10 (8) Mean rate per minute of total expressions, (SD)0.4 (0.1)0.4 (0.2)0.5 (0.3)Topic of uncertainty Diagnostic tests, mean % (SD)6 (7)6 (9)5 (9) Disease cause, mean % (SD)8 (9)4 (6)14 (14) Diagnostic label, mean % (SD)30 (15)39 (24)19 (19) Course, mean % (SD)25 (14)22 (16)25 (20) Medication, mean % (SD)31 (17)27 (18)44 (31)Initiative of expression Clinicians, mean % (SD)65 (16)66 (24)58 (27) Patients, mean % (SD)21 (12)27 (23)12 (19) Companion, mean % (SD)10 (10)8 (12)12 (18)Explicit expressions, mean % (SD)26 (15)24 (17)23 (21) Total communication approaches, N (SD) 14 (6) 12 (10) 23 (12) Information-focused approachesa, mean % (SD)72 (13)74 (16)69 (25) Give a reason for uncertainty, mean % (SD)20 (10)21 (13)16 (13) Alternate uncertainty with certainty, mean % (SD)18 (11)19 (14)15 (15) Outline potential scenarios and consequences, mean % (SD)9 (9)6 (9)15 (18) Comprehensibly explain uncertainty, mean % (SD)1 (2)0 (2)1 (5) Check understanding of the uncertain information, mean % (SD)0 (1)0 (1)0 (0) Offer sense of control, by reducing uncertainty, mean % (SD)25 (10)27 (17)23 (15)Emotion-focused approachesa, mean % (SD)15 (9)15 (12)13 (18) Warn for ongoing uncertainty, mean % (SD)2 (4)2 (3)2 (3) Explore patient's uncertainty tolerance, mean % (SD)0 (1)0 (1)0 (0) Offer coping strategies for uncertainty, mean % (SD)0 (1)0 (0)0 (1) Offer hope, mean % (SD)11 (9)12 (11)9 (17) Offer emotional support or space for emotional response), mean % (SD)1 (3)1 (3)1 (3) Emphasis on ongoing availability, mean % (SD)1 (2)1 (2)1 (3)Other approachesa, mean % (SD)13 (10)11 (10)18 (22) Tempering hope, mean % (SD)4 (7)4 (7)4 (9) Directing patients' perception, mean % (SD)6 (7)5 (7)7 (11) Other approaches, mean % (SD)2 (4)1 (4)7 (19)Patient and companion expressions of uncertaintyAll consultations (N = 50)First consultation (N = 50)Second consultation (N = 41) Total expressions, mean frequency (SD) 11 (8) 7 (5) 6 (4) Expressed by patients, mean percentage (SD)73 (25)75 (31)76 (28)Mean rate per minute of total expressions, (SD)0.2 (0.1)0.2 (0.2)0.2 (0.2)Clinician response to patient/companion uncertainty expression Give direct information, mean % (SD)67 (18)65 (25)62 (31) Explore, mean % (SD), mean % (SD)6 (8)7 (11)6 (11) Give indirect information, mean % (SD)2 (7)2 (8)3 (6) Give minimal recognition, mean % (SD)8 (9)10 (17)6 (11) Ignore, mean % (SD)7 (11)7 (17)12 (26) Reflect, mean % (SD)10 (14)10 (14)12 (24)aPercentages add up to >100% because maximally 2 communication approaches could be selected for each expression of uncertainty. Descriptive Statistics for Observed Uncertainty Communication During Recorded Consultations Percentages add up to >100% because maximally 2 communication approaches could be selected for each expression of uncertainty. Table 3Examples of Observed Communicative Approaches to Expressing UncertaintyInformation-focused approachesExample 1. Give a reason for uncertaintyActually, you couldn't confirm the diagnosis until you examined the brain under a microscope (topic: Preliminary diagnosis; implicit) 2. Alternate uncertainty with certainty or indicate most probable scenarioYou don't meet the criteria for diagnosing Parkinson's disease with sufficient certainty. But I do think that that's where it's heading, considering your symptoms (topic: Preliminary diagnosis; implicit) 3. Outline potential scenarios and consequencesSome people respond well for y to the same dosage of levodopa, while others need to increase their dosage within a few y. I can't predict this in advance (topic: Medication dosage; explicit) 4. Comprehensibly explain uncertaintyWe know that the diagnosis is incorrect in 10% of cases. So, 90% of the time, it is Parkinson's disease, while in the remaining 10%, it turns out to be something else (topic: Differential diagnosis; implicit) 5. Check understanding of the uncertain informationSo this [the patient's disease] is Parkinson's, but only detectible with a magnifying glass. That doesn't mean you'll necessarily be severely affected by it. Do you understand what I mean? (topic: Course of symptoms over time; implicit) 6. Offer sense of control, by reducing uncertaintyAnd imagine that we start you on the medication and the results are not particularly impressive. In that case, there's always the option of surgery (topic: Effect of medication on symptoms; implicit) Emotion- focused approaches 7. Warn for (ongoing) uncertaintyPeople simply respond variably to the medication. We'll just have to see how it works for you (topic: Effect of medication on symptoms; implicit) 8. Explore patients' uncertainty tolerance Not observed 9. Offer coping strategies for uncertaintyI think you shouldn't put your life plans off any longer. Because you never know what the future holds. No 1 has a crystal ball (topic: Course of symptoms over time; explicit) 10. Offer hopeI understand that this news is very intense. But I hope you also hear me say: during the first y there's a very good chance that you will start feeling much better (topic: Course of symptoms over time; implicit) 11. Offer emotional support or space for emotional responseYeah, so it is very difficult for me to say that I can't predict whether and when you will have difficulty with movements and other symptoms. So it is very disappointing that I cannot give you a clear “package”—like: this is what you should expect (topic: Course of symptoms over time; explicit) 12. Emphasize ongoing availabilityAnd of course you will further read about various things like care needs in the disease trajectory. That won't necessarily happen, but it's possible. Let's revisit this another time and feel free to ask me about it (topic: Course of symptoms over time; implicit) Other approaches 13. Temper hopeYou will read that your life expectancy is not shortened by the disease. I think it is shortened somewhat - but it's very difficult to predict on an individual basis. And nowadays there is a lot we can do with medication and deep brain stimulation (topic: Course of symptoms over time; explicit) 14. Implicit steering by minimizing/maximizing uncertaintyThe chances of finding a genetic cause for your disease are extremely low, only 3 to 7 percent. And even if we would find something, we would not act on it, so it's essentially meaningless (topic: Genetic predisposition; implicit) 15. OtherFor example, offering the patient a choice: We could do a DaTscan to tell whether your type of tremor better matches Parkinson's disease or an essential tremor. We could also decide to let it run its course or treat the tremor itself, and then we can see what the future brings. But I don't know how you would feel about this? (topic: Desirability of diagnostic test; implicit) 16. No specific approach observedWe'll need to explore which dosage is best suited for you (topic: Medication dosage; explicit) Examples of Observed Communicative Approaches to Expressing Uncertainty Patients and companions expressed uncertainty on average 11 times over both consultations combined (SD = 8, range 1–29), at a rate of 0.2 p/m (SD = 0.1, range 0–0.7). Of these expressions, 73% were made by the patient (SD = 25, range 15–100). Clinicians most often responded to patient-expressed or companion-expressed uncertainty by providing more information (67%), for example, “Patient: “You hear so often about people who fall. Is that …?” Clinician: “Yes, that is part of the disease.” Changes in Patient-Reported Outcomes and Effects of Uncertainty Tolerance Before the first consultation, patients' uncertainty was not correlated with their age, sex, or educational level. Greenhouse-Geisser–corrected repeated-measures ANOVA indicated changes in patients' uncertainty feelings over time (F(2.4, 76.8) = 13.4, p < 0.001, η2 = 0.30; Figure 3). Specifically, uncertainty was decreased at T1 compared with T0 (mean diff. = −0.6, p < 0.001, d = 0.77). At T2 and T3, levels of uncertainty did not further change (both not significant). Concerning patients' trust, repeated-measures ANOVA indicated no changes over time (F(2, 66) = 2.3, p = 0.11, η2 = 0.07; Table 4). Regression analyses indicated that clinicians' uncertainty communication (frequency, explicitness, or emotion-focused approach) did not predict changes in patients' feelings of uncertainty or trust in their clinician (Table 4). Moreover, these effects were not significantly moderated by patients' tolerance of uncertainty. Linear mixed-effects regression models indicated that clinician uncertainty tolerance did not predict their uncertainty communication (frequency, explicitness, or emotion-focused approach; all p values not significant; eTable 1). Patients' Feelings of Uncertainty and Trust in Their Clinician Over Time (A) T0: 0–2 weeks before first consultation. Feelings of uncertainty were assessed for n = 49 patients. (B) T1: 0–2 after first consultation. Feelings of uncertainty and trust in clinician were assessed for n = 46 patients. (C) T2: 0–2 weeks after second consultation. Feelings of uncertainty and trust in clinician were assessed for n = 41 and n = 40 patients, respectively. (D) T3: 6–8 weeks after second consultation. Feelings of uncertainty and trust in clinician were assessed for n = 44 and n = 43 patients, respectively. (E) Repeated-measures ANOVA results for n = 33 patients (feelings of uncertainty) and for n = 34 patients (trust in clinician). (F) Potential range = 1–5. Higher scores indicate higher uncertainty. (G) Potential range = 1–5. Higher scores indicate higher levels of trust. ANOVA = analysis of variance. Table 4Effects of Clinician Uncertainty Communication on Patient Trust and Changes in Uncertainty Over Time in Multiple Regression AnalysesFeelings of uncertaintyTrust in the clinicianChange in patients' feelings of uncertainty at T0-T1aTrust at T1cbSE Bβp ValuebSE Bβp ValueConstant−0.400.350.334.240.21<0.001Number of uncertainty expressions during first consultation−0.020.03−0.100.56−0.000.02−0.040.81Explicit expressions (%)−0.000.01−0.050.760.000.00−0.010.98Emotion-focused communication approaches (%)−0.000.01−0.030.850.000.01−0.000.98Change in patients' feelings of uncertainty at T1-2bTrust at T1b,dConstant−0.370.210.09−0.180.190.37Number of uncertainty expressions during a consultation−0.000.02−0.050.820.010.020.110.61Explicit expressions (%)0.000.010.190.32−0.000.01−0.040.84Emotion-focused communication approaches (%)0.010.010.160.440.010.010.140.51aR2 = 0.02.bR2 = 0.54.cR2 = 0.00.dR2 = 0.05. Effects of Clinician Uncertainty Communication on Patient Trust and Changes in Uncertainty Over Time in Multiple Regression Analyses R2 = 0.02. R2 = 0.54. R2 = 0.00. R2 = 0.05. Discussion We investigated how uncertainty is discussed during diagnostic consultations about PD, the potential effects of such communication on patients, and the role of patients' and clinicians' uncertainty tolerance. We found that uncertainty—particularly regarding the diagnostic label, disease course, and medication effects—was integral to communication, yet often discussed implicitly. Clinicians focused on reducing their own or patients' expressed uncertainty, for example, by providing additional information or follow-up plans. Fewer efforts were observed to address the emotional impact of uncertainty and how patients could manage it. Patients' levels of experienced uncertainty were moderate, on average, and decreased significantly over time—particularly directly after their first diagnostic consultation. Changes in patients' feelings of uncertainty or trust were not explained by how clinicians discussed uncertainty. Finally, we found no evidence of any moderating role of patient or clinician uncertainty tolerance. Uncertainty is inherent to a PD diagnosis,6 which was clearly reflected in the frequency of observed uncertainty expressions. Although previous research implied that clinicians sharing uncertainty might negatively affect patients, we observed no such impact on patients' feelings of uncertainty or trust.22,23 Potentially, instead of frequency, the level of explicitness (e.g., “I don't know” vs “it is possible that”) or communicative approach may determine how patients are affected by uncertainty.43 Several other explanations might account for the lack of effects of uncertainty communication on patients' well-being observed. Patients' feelings of uncertainty reported shortly after diagnosis may be primarily affected by the inherently uncertain nature of the diagnosis itself, thereby overriding any potential impact of clinicians' ways of communicating about the associated uncertainty. Alternatively, patients' responses to clinicians' uncertainty communication may be variable, resulting in a net lack of effect. For example, some patients may focus on hopeful scenarios when confronted with prognostic uncertainty (“perhaps disease progression will be slow in my case”), whereas people envisioning pessimistic scenarios might experience more negative emotions. Regarding patients' trust, similar variation between patients might explain why no overall effects of uncertainty communication were observed. For example, some patients' trust may be eroded when clinicians acknowledge uncertainty, whereas others could perceive it as evidence of the clinician's honesty.43 Alternatively, patients' trust may have remained unaffected because it was assessed so briefly after diagnosis. Previous research suggests that solid and substantiated trust only develops after repeated interactions with a health care provider.44 Clinicians' communication contained mainly uncertainties regarding the diagnostic label in the first diagnostic consultation and regarding medication in the second. Of interest, clinicians also relatively frequently expressed uncertainty about the disease course. This may seem surprising because, in a separate analysis of these same consultations, we found that information provided about the disease course was scarce.29 This discrepancy suggests that almost any information clinicians provide about patients' disease course may be inherently uncertain. After their consultations, 30% of patients reported a need for more information regarding the future implications of PD, suggesting that patients' prognostic information needs could not be met.11,29 Clinicians may feel challenged to address patients' need for more prognostic certainty, whereas prognosis is inherently unpredictable in PD. We observed some helpful examples of how clinicians could openly acknowledge prognostic uncertainty, while also providing some sense of control or hope—for example, by sketching the best case, worst case, and most likely scenario. In addition, clinicians may explicitly stress the impossibility of prognosticating individual patients' disease course, to encourage patients in accepting that their disease is inherently unpredictable. We found that clinicians expressed uncertainty mostly subtly, using implicit phrasing (e.g., using adverbs such as “possibly”), which aligns with previous research in other settings.34 Evidence suggests that, at the short term, patients may prefer and even psychologically benefit from such implicit expressions of uncertainty.22 As symptoms worsen at the longer term, however, patients may need a more conscious awareness of uncertainty to cope with their disease. Clinicians' explicit communication about these uncertainties at an early stage may serve to manage patients' expectations and facilitate their disease management.45 Reassuringly, we found no indication that clinicians' explicit communication of uncertainty increases uncertainty or decreases trust. Thus, patients may be able and willing to accept and deal with uncertainty even when explicitly confronted with it. Nevertheless, patients may vary in their ability to process explicit uncertainty at diagnosis. For example, patients who already suspect that they may have PD might be better able to process and manage the explicit uncertainties associated with the diagnosis compared with unprepared patients. Neurologists may tailor their level of explicitness and the overall amount of uncertain information to patients, based on their perception of each patient's ability to deal with uncertainty. If clinicians sense that patients are overwhelmed, or if they need more time to explain complex uncertainties, clinicians could monitor patients' openness to postponing further information to subsequent consultations. Clinicians could further enhance patients' acceptance of uncertainty by addressing its psychological impact. In our data, clinicians focused mostly on the informational aspects of uncertainty. For example, clinicians accounted for their own uncertainty and provided additional information or diagnostic tests to reduce uncertainty. Such information-focused approaches may effectively enhance patients' understanding of uncertainty and offer the prospect of more certainty.21 Yet, many uncertainties regarding PD (e.g., whether the diagnosis is accurate) cannot be eliminated or reduced at diagnosis. For such inevitable uncertainty, supporting patients in dealing with its impact might be helpful.17 For example, after openly acknowledging uncertainty, clinicians might explore patients' experiences and suggest corresponding coping strategies.46 In our study, we occasionally observed how clinicians would acknowledge the emotional impact of ongoing uncertainty, for example, by offering hope or emphasizing their ongoing availability for the patient. Communication training might further facilitate clinicians in combining information with emotion-focused approaches when discussing uncertainty. Clinicians' uncertainty tolerance did not predict their uncertainty communication, nor was patients' uncertainty tolerance related to how they were affected by uncertain information. Potentially, other factors, such as the amount of perceived social and clinician support, may more strongly determine how patients with PD are affected by uncertainty than uncertainty tolerance.47 Alternatively, this finding may be explained by limitations in the uncertainty tolerance measures used in our study. Existing measurements may approach “tolerance” too simplistically, with high uncertainty tolerance as a desirable trait or state, whereas, in reality, both extreme tolerance and extreme intolerance may be potentially harmful.48 More fine-grained insights are necessary into the multiple and complex ways in which people may respond to uncertainty.49 Such nuanced understanding and assessment will eventually enable a better understanding of how uncertainty tolerance relates to uncertainty communication. Strengths of this work include our access to recorded consultations, which enabled us to analyze communication as it unfolds in real life. This yielded novel insights compared with existing related literature, which used mainly self-report. Moreover, this study longitudinally mapped how patients' uncertainty evolves throughout the diagnostic process. Thus, we were able to combine detailed insights into communication with patient self-report, enabling testing associations between observed interaction and subjective experiences. The inclusion of consultations conducted by 28 different clinicians in 6 different hospitals across the Netherlands reflects the diversity of clinical practice regarding PD diagnosis disclosure in the Netherlands. A first limitation is that systematically coding uncertainty communication is challenging. Particularly, in patients' communication, the distinction between a simple question and an actual expression of uncertainty was difficult to define. As a result, inter-rater reliability was limited, necessitating double-coding and discussion of all systematic coding. Although this approach ensured reliability for this study, future work needs to further refine this coding measure. Second, participating clinicians may have (un)consciously modified their communication behavior because of the camera presence. Although they reported minimal hindrance, this may have led to an overly positive impression of their regular practice. Third, our sample size was relatively small. To detect more complex interactions, future studies with larger samples would be necessary. Fourth, our patient sample was not maximally representative of the PD patient population. Specifically, we only included patients from the Netherlands who understood the Dutch language. In addition, many clinicians included in our sample were to some degree specialized in movement disorders. Thus, our results cannot be simply generalized. Future research might specifically explore how uncertainty is discussed in medical consultations involving language and cultural barriers, and in other geographic and cultural settings. In addition, future research could examine whether patients perceive uncertainty differently when conveyed by movement disorder specialists compared with generalist neurologists. Our results may reassure clinicians that conveying uncertainty seems not to be harmful for patients with PD, even when done explicitly. After the diagnostic consultations, in which uncertainty was frequently expressed, patients' feelings of uncertainty were still reduced overall. This suggests that clinicians can be transparent about limited available knowledge regarding certain PD topics. This finding may be reassuring, in light of a general hesitance in medicine to convey uncertainty, for fear it may reduce patients' trust or clinicians' reputation.19,20,50 Future research should establish whether this finding is generalizable to other diagnostic consultations and conditions. Clinicians may be supported not only to openly share uncertainty with patients and/or aim to reduce it, but also to facilitate patients' acceptance and management of uncertainty. Clinicians could, for example, use meta-communication, by acknowledging that sustained uncertainty may be difficult to manage and exploring a patient's response to it (for example, “Many people find the enduring uncertainty associated with a Parkinson's disease diagnosis difficult. What is that like for you?”). Based on their clinical expertise and experiences with other patients, clinicians could suggest coping strategies to make the uncertainty more bearable (for example, “Some people like to discuss their uncertain situation with friends and families, while others prefer not thinking about it too much. What would help you in making the uncertainty more manageable?”). In addition, they could jointly outline follow-up steps (for example, “Here is what we will do to try to get more certainty about your diagnosis”). If patients seem unable to manage uncertainty, clinicians could explore options for referral to psychological support. During PD diagnostic consultations, clinicians provide patients with extensive information about the disease, each element of which entails uncertainties. Findings of our observational study suggest that uncertainty is an integral part of these interactions, as evidenced by frequent expressions by clinicians, patients, and patients' companions. Yet, its discussion often remains implicit and is strongly focused on justifying or reducing the uncertainty. Our finding that uncertainty communication, even when explicit, was unrelated to patients' feelings of uncertainty or trust may reassure clinicians. This suggests that such communication does not necessarily harm patients' psychological well-being. Future research using larger samples and/or experimental designs could more systematically assess the effects of various ways to convey uncertainty on patients' well-being, understanding, and trust. Clinician training may facilitate them to more explicitly acknowledge uncertainty while also facilitating patients' acceptance and management of irreducible uncertainties.TAKE-HOME POINTS→ Clinicians, patients, and patients' companions frequently express uncertainty during diagnostic consultations in Parkinson disease (PD), emphasizing that uncertainty is integral to PD diagnosis.→ Clinicians' explicit communication about uncertainty did not seem to harm patients psychologically, suggesting that clinicians can be transparent about limited available knowledge.→ Discussions of uncertainty often remained implicit and were strongly focused on justifying or reducing the uncertainty.→ Clinicians may be supported through communication training to further facilitate patients' acceptance of and coping with uncertainty.→ Future research using larger samples and/or experimental designs could more systematically assess the effects of various ways to convey uncertainty on patients. TAKE-HOME POINTS → Clinicians, patients, and patients' companions frequently express uncertainty during diagnostic consultations in Parkinson disease (PD), emphasizing that uncertainty is integral to PD diagnosis.→ Clinicians' explicit communication about uncertainty did not seem to harm patients psychologically, suggesting that clinicians can be transparent about limited available knowledge.→ Discussions of uncertainty often remained implicit and were strongly focused on justifying or reducing the uncertainty.→ Clinicians may be supported through communication training to further facilitate patients' acceptance of and coping with uncertainty.→ Future research using larger samples and/or experimental designs could more systematically assess the effects of various ways to convey uncertainty on patients. → Clinicians, patients, and patients' companions frequently express uncertainty during diagnostic consultations in Parkinson disease (PD), emphasizing that uncertainty is integral to PD diagnosis. → Clinicians, patients, and patients' companions frequently express uncertainty during diagnostic consultations in Parkinson disease (PD), emphasizing that uncertainty is integral to PD diagnosis. → Clinicians' explicit communication about uncertainty did not seem to harm patients psychologically, suggesting that clinicians can be transparent about limited available knowledge. → Clinicians' explicit communication about uncertainty did not seem to harm patients psychologically, suggesting that clinicians can be transparent about limited available knowledge. → Discussions of uncertainty often remained implicit and were strongly focused on justifying or reducing the uncertainty. → Discussions of uncertainty often remained implicit and were strongly focused on justifying or reducing the uncertainty. → Clinicians may be supported through communication training to further facilitate patients' acceptance of and coping with uncertainty. → Clinicians may be supported through communication training to further facilitate patients' acceptance of and coping with uncertainty. → Future research using larger samples and/or experimental designs could more systematically assess the effects of various ways to convey uncertainty on patients. → Future research using larger samples and/or experimental designs could more systematically assess the effects of various ways to convey uncertainty on patients.